ASSOCIATIVE ADJUSTMENTS TO REDUCE ERRORS IN DOCUMENT SCREENING

Abstract

Associative adjustments to a document file were considered as a means for improving retrieval. The investigation includes the definition and theoretical investigation of the statistical properties of a generalized mismatch measure. Improvements in retrieval resulting from performing associative regression adjustments on data file are examined both from the theoretical and experimental point of view. The expected gain in mismatch is presented as a function of various measurable characteristics of the file, such as error rates in indexing and the probability distributions of the associative adjustment criteria. Query adjustments using negative as well as positive correlations are considered and found to be ineffective. In a limited site Patent Office file with a low indexing error rate experimental results are presented applying (1) no associative correction (2) the generalized mismatch with no associative correction (3) associative correction and (4) query adjustment. In general the results using the ordinary mismatch with an associative adjustment are superior to those using the more generalized quadratic mismatch or the query adjustment scheme.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 31, 1967
Accession Number
AD0651630

Entities

People

  • David G. Weinman
  • Donald T. Searls
  • Edward C. Bryant
  • Robert H. Shumway

Organizations

  • Westat

Tags

Communities of Interest

  • Energy and Power Technologies
  • Human Systems
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Air Force
  • Classification
  • Computer Programming
  • Computers
  • Covariance
  • Frequency
  • Index Terms
  • Indexes
  • Information Retrieval
  • Information Science
  • Normal Distribution
  • Patent Office
  • Probabilistic Models
  • Probability
  • Probability Distributions
  • Random Variables
  • Scientific Research

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Computational Modeling and Simulation
  • Parallel and Distributed Computing.